Traffic Signal Timing Scheme Recommendation System Design And Analysis

With the advancement of urbanization in China,traffic congestion has become the main issue for large and medium-sized cities.Urban road intersections are important nodes that integrate traffic flow in all directions and are at a critical position in the urban road network system.Therefore,signal timing control at road intersections has become a core issue in managing urban traffic congestion.The traditional signal timing control required a lot of manul time,and the effect of the program implementation depends on the past experience of the specialists,which is difficult to meet the needs of synchronous regulation.Therefore,it is of great significance to study a traffic signal timing scheme recommendation system in order to perform efficient and intelligent traffic control.This research studies the designs on similar models of the intersections and the recommended algorithms.By introducing the intersection similarity,a customized recommendation algorithm built on the large data suitable for traffic signal timing distribution is constructed.Meanwhile,the overall architecture design of a new system is carried out based on the functionality and performance by summarizing the current issues of the traffic signal control system and the requests from specialists on the signal timing distribution via surveys.Basic information configuration function of the intersection is achieved through the structural storage of data from intersections.The properties of intersections are retrieved and distributed based on the frequency of occurrence on the ground of research on TF-IDF(Term Frequency–Inverse Document Frequency)algorithm and statistical theory.Considering the dynamics of the traffic system and the particularity across times and spaces,the intersection static similarity model and the intersection dynamic similarity model are established respectively.The first step of filtering isperformed by using a user-based collaborative filtering recommendation algorithm to introduce static similarity of intersections.The project-based collaborative filtering recommendation algorithm is used to introduce the intersection dynamic similarity,and the final recommendation result is generated,and the task of recommending the timing scheme to the signal timing personnel is completed.User management,intersection classification,congestion alarm and other functions are also designed as a whole in this study.The reliability and the validity of the proposed algorithm is verified with the experiments.